基于双线策略的高亏格三维曲面传感器网络的分布式数据存储与查询机制研究

In recent years, frequently occurred accidents such as high-rise building fires, coal-mine explosions and petroleum pipeline leakages, cause heightened concerns on a kind of wireless sensor networks (WSNs) where sensors are typically deployed in complex-connected 3D surfaces. This kind of WSNs is often of a complex-connected 3D setting and has non-trivial topology, possibly with high genus (i.e., multiple handles), and thus is called high genus 3D surface WSNs. In this research, we focus on distributed data storage and retrieval scheme in high genus 3D surface WSNs. Our scheme is based on duble rulings method. We first reveals the properties of connectivity, closedness, as well as their universal conditions of the iso-contours of high genus 3D surface WSNs based on the continuous geodesic patterns, where the data storage curve is constructed on the basis of the iso-contours. Then the theories of topology identification and computational geometry are introduced to achieve genus identification based network decomposition, using geometry features of the network. On this basis, the data retrieval curve is designed so as to ensure the load balance of data storage, the distance sensitivity of data retrieval, and the locality of node operation. This research will help to get better understandings and bring new views of the geometry features of high genus 3D surface WSNs; it will also provide important technical reserve and theoretical foundations for the development of high genus 3D surface monitoring and early-warning networks.

近年来频繁发生的高层楼宇火灾、煤矿爆炸、输油管泄漏等事故引起了研究人员对一类布放在三维空间表面的传感器网络的重视。这类传感器网络由于具有较为复杂的拓扑，且通常有多个亏格（即“孔洞”），因而被称为高亏格三维曲面传感器网络。本项目主要研究高亏格三维曲面传感器网络的分布式数据存储与查询机制。拟基于双线存储与查询策略，首先揭示基于连续测地曲线的高亏格曲面网络的等值线所具有的连通性、封闭性及其普适条件，并以等值线为基础构建网络的数据存储曲线；然后引入拓扑识别和计算几何的相关理论，利用网络的几何特征，实现基于亏格识别的网络分解，并在此基础上设计网络的数据查询曲线，从而保证数据存储的负载均衡性、数据查询的距离敏感性和节点操作的局部性。本项目的研究有助于加深对高亏格曲面网络几何特征的理解，带来观念上的新认识；同时也为推动高亏格三维曲面传感器监测预警网络的研发提供重要的技术储备和理论依据。

无线传感器网络中的许多应用要求在给定监测区域中的传感器观测数据以串行的方式聚合。这需要构建穿过在该领域的所有传感器的路由路径，同时也需要线性化网络。在本项目中，我们介绍了SURF——一个高亏格三维曲面的无线传感网络的空间填充曲线构建方案，产生非周期性的遍历路径（也就是说，任何节点最多被覆盖一定次数）。SURF首先利用最大跳跃距离函数在离散设置中构造轮廓，然后使用概念Reeb图和最大割集将网络划分到不同的区域。最后，它进行一种新颖的串行遍历方案以使能够在区域之间和之间遍历。据我们所知，SURF是第一个以高亏格三维曲面无线传感网络为目标的和纯基于连通性的线性化网络算法。它具有完全分布性和高度可扩展性，且只需要恒定的存储空间以及网络中每个节点的通信成本。为了结合适应密度的构造空间填充曲线，我们还设计了第二个算法，称为SURF+。它利用参数化的螺旋状曲线覆盖三维表面，从而产生适应不同旅行预算或融合延迟要求的多分辨率SFC。在此基础上，我们提出了结合了网络内数据存储算法和高亏格三维曲面无线传感网络中的检索算法的应用程序。对几个有代表性的网络广泛仿真表明所提出的算法在高亏格三维曲面无线传感器网络上都具有很好的性能。